Evaluation Performance Neural Network Genetic Algorithm

DOI

10.2991/aisr.k.200424.065How to use a DOI?

Keywords

feed forward neural network, evaluation, rainfall

Abstract

Hybrid models of the precipitation spillover process that are embedded with high unpredictability, non-stationarity, and non-linearity inboth spatial and worldly scales can give significant results in rainfall forecasting. Considering this, various neural network models have been applied to reproduce this complex process. Neural Network is an information processing system that has characteristics similar to biological terms. A neural network is a machine designed to model the workings of the human brain in performing certain functions or tasks. In general, FFNNs are trained to use the Backpropagation algorithm to get their weights. Backpropagation can work well on simple training problems, but its performance will decrease and be trapped in a local minimum when applied to data that has enormous complexity. Therefore, metaheuristic operations are needed using Genetic Algorithms (AG). In this paper, a detailed discussion of the FFNN-AG step construction will be given, which is a search algorithm based on selection and genetic mechanisms to determine global optimum and evaluation of previous paper related to this.

Copyright

© 2020, the Authors. Published by Atlantis Press.

Open Access

This is an open access article distributed under the CC BY-NC license (http://creativecommons.org/licenses/by-nc/4.0/).

Download article (PDF)

TY  - CONF
AU  - Hetty ROHAYANI
AU  - Tuga MAURITSIUS
AU  - Leslie H Spit Warnars Harco
AU  - Edi ABDURRACHMAN
PY  - 2020
DA  - 2020/05/06
TI  - Evaluation Performance Neural Network Genetic Algorithm
BT  - Proceedings of the Sriwijaya International Conference on Information Technology and Its Applications (SICONIAN 2019)
PB  - Atlantis Press
SP  - 426
EP  - 431
SN  - 1951-6851
UR  - https://doi.org/10.2991/aisr.k.200424.065
DO  - 10.2991/aisr.k.200424.065
ID  - ROHAYANI2020
ER  -